[[email protected]]

Hello again. I'm still working on my tube testing project. I have been
trying to determine the transconductance of the 6EJ7 tubes I have. Here is
my setup: Plate and screen grid at 240 vdc. Heater at 6.3 vdc. Grid 3
tied to cathode. Control grid at -3 vdc tied to cathode through 10 k
resistor. Plate current is reading normal and when I vary the grid voltage
and devide the change in plate current by the change in control grid voltage
and multiply by 10 to the 6, I get the expected transconductance. I have
done this using the 240 vdc plate curve and the 200 vdc plate curve. When I
adjust the control grid voltage from -3 volts towards -2.5 volts the plate
current starts running away and the control grid can't be adjusted any
further. I have to turn the plate supply off to regain control. Using 200
vdc as the plate voltage and approximately -2.75 volts on the control grid I
measure 15 ma and riseing. Is this normal or is this indicative of a
problem?
Thanks,
D. E.

[Patrick Turner]

wrote:

Hello again. I'm still working on my tube testing project. I have been
trying to determine the transconductance of the 6EJ7 tubes I have. Here is
my setup: Plate and screen grid at 240 vdc. Heater at 6.3 vdc. Grid 3
tied to cathode. Control grid at -3 vdc tied to cathode through 10 k
resistor. Plate current is reading normal and when I vary the grid voltage
and devide the change in plate current by the change in control grid voltage
and multiply by 10 to the 6, I get the expected transconductance. I have
done this using the 240 vdc plate curve and the 200 vdc plate curve. When I
adjust the control grid voltage from -3 volts towards -2.5 volts the plate
current starts running away and the control grid can't be adjusted any
further. I have to turn the plate supply off to regain control. Using 200
vdc as the plate voltage and approximately -2.75 volts on the control grid I
measure 15 ma and riseing. Is this normal or is this indicative of a
problem?
Thanks,
D. E.


I assume you have the set up as follows...

Anode and screen tied together AND both taken to a low impedance 240Vdc
supply.

Cathode to 0V.

Grid 3 taken to 0V,

Grid 1 taken to a low impedance variable voltage supply centred around
-3Vdc.

Is there a 10k resistance between the grid 1 and its variable biasing
Vdc supply?

If the grid 1 voltage is made mode positive say to -2.5V,
and you say that Ia "runs away", ie after a short time the Ia becomes
much greater than
the gm should make it, then what voltage is across the series 10k g1 R?

What is screen current in relation to anode current at g1 = -3V?

Is it anode current or screen current that "runs away"?

If there is a 10k R between Vdc g1 bias supply, does the Eg1 go more
positive than the bias supply?

Ie, does a largish Vdc appear acros the series R to g1?

Is the tube oscillating at some RF ?

Is the anode and screen supply well bypassed with C placed close to A
and g2
and cathode/0V?


I am sure had you more thouroghly measured everything properly and more
quickly,
and tried several samples of this tube, that you'd have found all the
answers your'e looking for.

If gm was around 10mA/V, expect Ia to increase 5mA with a +0.5V change
at g1.

What happens to the screen current?

Looking closely with magnifying glass at the tube, does the screen
appear to glow
red/orange?

Data for the tube can be seen at
http://tdsl.duncanamps.com/pdf/ef184.pdf

The tests shown include Ra curves, ie, Ia vs Ea curves which indicate Ra
for various
Eg1 values AND on 3 graphs, with Eg2 = +140V, 170V, and 200V, but not
240V.

The data says that with Ea = Eg2 = 230Vdc, Ia = 10mA, Ig2 = 4.1mA, and
Eg1 would probably have to be -3V or more.
So Pda for anode = 2.3W, and PdG2 = 0.943W.

What did you measure?

What currents have you observed?

To measure the currents, two 100 ohms can be used as sensing R in series
with g2 and A, without affecting results much.

( 5mAdc change across 100 ohms = 0.5Vdc only ).

Try several different samples to make sure you have not got a dud tube,
because your experience
looks like you have got a dud.

You can of course use ac to test the gm.

If 1Vac ( Vrms ) at say 1kHz from a sig gen is applied to g1 via a
coupling cap
and low value Rg, then if gm was 10mA/V, expect to see 10mArms of Ia at
the 100 ohms, ie, 1Vrms.
Your CRO should show slight 2H distortion, because the load of 100 ohms
is near enough to a vertically
drawn line on the anode curves.
Ac Screen current will also be seen, and you should measure and observe
its magnitude
and distortion nature.

Patrick Turner.

[[email protected]]

Thanks Patrick. I need to review this a little. I am away from the circuit
at this time and can't do much with it.. I will review what you hasve
provided and apply Monday. I will post back Monday evening with results.

Thanks again,
Don Ennis

[[email protected]]

I had a little more time to look at the circuit today and make some
measurements. I reset the plate and screen grid to 200 vdc. I varied the
g1 voltage from -5 to -2.4 and recorded Ib at normal levels. Ic2 also
seemed normal. I have g1 connected to cathode through the 10k resistor. I
need to look at a few more things as I am told that these tubes were not
outputting up to spec. I can't recreate the circuit they were used in so
I'm just doing the best I can to verify they have low output. So far they
seem ok. Any thoughts?

Thanks,
Don